Seismological Society of America Annual Meeting Fieldtrip 2015

Seismological Society of America Annual Meeting Fieldtrip 2015 Faults, vistas, and earthquakes in the northern Los Angeles and Pasadena Region April 24, 2015 Kate Scharer (USGS), James Dolan (USC), Jerry Treiman and Janis Hernandez (CGS) and Doug Yule (CSUN)

Route (black line), stops (black stars), on map adapted from Geologic Map of Los Angeles Quad (Yerkes and Campbell, 2005). Main faults discussed on the trip (red lines) are generalized. Accurate information on fault location and activity can be found on the CGS and USGS websites. CGS: http://www.consrv.ca.gov/cgs/rghm/ap/Pages/main.aspx USGS: http://earthquake.usgs.gov/hazards/qfaults/

THE PLAN 8:00-8:15 Load and depart! 9:00-10:30 Griffith Observatory (Dolan) 10:30-12:00 Hollywood Blvd/Argyle St (Treiman, Hernandez) 12:00-1:15 Wattles Park for lunch (Dolan) 1:15-1:45 Autry Overlook 2:15-4:00 Northridge Recreation Center and CSU Northridge (Yule) 4:00-5:00 Loma Alta Park (Scharer, Dolan)

START: Pasadena Convention Center

Route travels west out of Pasadena on Highway 134. Accessing the highway, the bus will cross Arroyo Seco, a drainage that forms the western border of Pasadena (the Rose Bowl is about a mile to our north). The highway cuts the southern tip of the Verdugo Mountains, along the Verdugo-Eagle Rock Fault system, a northeast-dipping, northwest-trending fault with estimated < 1 mm/yr slip rate. No neotectonic studies with slip rates exist for Verdugo Fault, but Weber et al., (1980) report south facing scarps in Holocene alluvium, and deformation models by Cooke and Marshall (2006) predict it is oblique, with ~0.6 mm/yr right-lateral and 0.5 mm/yr dip slip rates. The Verdugo Mountains are largely Cretaceous gneissic quartz diorite; low terrain to the south is Miocene Topanga Group conglomerates and siltstones (Yerkes and Campbell, 2005).

Head south on Interstate 5. Here Interstate 5 parallels the Los Angeles River for about 4 km on the western edge of the Santa Monica Mountains; at this location they are known as the Hollywood Hills. Rocks immediately west of I-5 are Miocene Monterey Formation shale; these overlie the granodiorite basement rocks exposed to west at Griffith Observatory (Dibblee and Ehrenspeck, 1991).

 This low oblique Google Earth view looks westward along Los Feliz Blvd. in our direction of travel. An aerial photo of the area from 1927 is draped over the landscape (vertical exaggeration x2). Our route follows this eroded and breached fault-line trough along which remnants of the old topography are still visible. The magenta lines indicate interpreted fault traces. (1927 photo by Fairchild Aerial Surveys).

 Composite map of the Burbank (1926) and Glendale (1928) 6-minute topographic maps that show the Los Feliz Blvd. portion of our route. This map series nicely depicts the topography with 5-foot contours.

STOP 1: GRIFFITH OBSERVATORY

Our first stop is at the famed Griffith Park Observatory, notable for the nice view of the Hollywood sign atop the westernmost end of the Elysian Park anticlinorium (hanging wall fold above Elysian Park blind thrust fault), and for providing the backdrop for numerous movies, including perhaps most notably Rebel Without a Cause with James Dean (sorry, Charlie's Angels: Full Throttle!). The observatory is an ideal spot to introduce and discuss LA's urban fault network, as it provides expansive views of many of these structures.

At this stop we will discuss what is known about the major fault systems that underlie metropolitan Los Angeles, with a particular focus on the Sierra Madre-Cucamonga fault system, the Raymond-Hollywood- Santa Monica fault system, and the Puente Hills and Compton blind thrust faults, the two largest active blind thrust systems that underlie most of the area you can see from Griffith Park.

 Map of major faults in southern California from the Southern California Earthquake Center’s Community Fault Model (CFM). San Andreas fault (red) and Puente Hills and Compton blind thrust faults (orange) are highlighted. Thin red liens show surface traces of faults in the CFM. White lines denote coastline and California State boundaries. Image created by Andreas Plesch and John Shaw (Harvard University).

As we look southward over what is known as the LA basin, in addition to noticing the density of development, including the downtown LA high-rise district (which keeps rising higher every year), it is important to note that the "LA basin" isn't just a geographic term used by traffic reporters. Rather, the Los Angeles basin is also a deep (10 km at its deepest point), narrow, 50-km-long sedimentary trough that extends northwest-southeast beneath much of the urban core of LA. You can't see it because it has been filled with sediments eroded off the San Gabriel Mountains and deposited on the floodplains of the Los Angeles and San Gabriel Rivers. The LA basin, which has been the subject of extensive modeling efforts to understand the seismic hazard ramifications, will trap and amplify seismic waves during future earthquakes. Of particular importance in this regard is the proximity of Puente Hills and Compton blind thrust faults to the basin. Indeed, the location and geometry of the PHT render it perhaps a "worst-case- scenario" fault for metropolitan LA, as any upward rupture directivity will funnel energy directly into the downtown region and the deep sedimentary basin.

 Map of major blind thrust faults in the Los Angeles basin. Note segmented nature of Puente Hills blind thrust fault (orange) and Compton blind thrust ramp (yellow) extending beneath much of much of the western LA metropolitan area. Both blind thrust ramps dip gently northward (Shaw et al., 2002; Leon et al., 2009).

 3D cutaway image showing geometry of the central, Santa Fe Springs segment of the Puente Hills blind thrust fault through the site of the 1987 Mw 6.0 Whittier Narrows focus. Geometry of the uppermost 7 km is constrained by high-quality petroleum- industry seismic reflection data, whereas the deeper fault geometry is constrained by the location of the 1987 mainshock focus and relocated aftershocks. Data from Shaw and Shearer (1999) and Shaw et al. (2002).

 Northeastward-looking oblique view of depth to basement surface highlighting the presence of the deep Los Angeles basin (darker colors). Orange surfaces are the three main structural segments of the Puente Hills blind thrust fault. Blue surface to west of PHT is the CFM representation of the Newport- Inglewood fault, and the blue surface to the east is the CFM representation of the Whittier fault extending along the eastern end of the PHT. Red lines show surface traces of SCEC CFM faults, and the white line denotes coastline. Image created by Andreas Plesch and John Shaw (Harvard University).

NEXT: STOP 2: HOLLYWOOD BLVD/ARGYLE STOP (PANTAGES THEATRE)

Exit Griffith Observatory, south onto Vermont Ave, right onto Los Feliz Boulevard, left onto Western Ave, then right at Hollywood Boulevard. From Griffith Park we will drop down the southern side of the Hollywood Hills, and drive along Hollywood Boulevard. The saddle in the hill occupied by the 101 Freeway exposes marine clastic rocks of the Miocene Upper Topanga Formation.

 This vintage oblique aerial photo from 1938 is looking northeast across an elevated alluvial fan, north of Los Feliz Blvd. There are at least two fault scarps visible across this fan (yellow annotations). The area was developed by mass grading in 1962, before faulting was well understood or considered active in this area. (photo by Fairchild Aerial Surveys).

The south-flowing drainage at Fern Dell Drive appears to have been offset at least 130 meters in a left-lateral sense (locality marked S4e in this figure) with a younger fan starting to develop where Fern Dell crosses Los Feliz. (Figure modified from Figure 14 of Hernandez & Treiman, 2014).

Seismological Society of America Field Trip Stop – Hollywood/Argyle April 24, 2015

Fault Rupture Hazard Hollywood Fault Zone, Hollywood, California Jerry Treiman, Senior Engineering Geologist Janis Hernandez, Engineering Geologist California Geological Survey

This stop will provide an overview of fault rupture hazard recognition in the Hollywood area under the Alquist-Priolo Earthquake Fault Zoning Act (the "Act"). The Act was passed in 1972 as a direct result of surface rupture damage to buildings in the 1971 M 6.6 San Fernando Earthquake, damage that led to the recognition that surface fault rupture is a readily avoidable hazard. The intent of the Act is to reduce the risk from surface fault rupture by prohibiting the location of developments and structures for human occupancy across the trace of active faults. This goal is achieved through a division of responsibilities between the State, local government and landowners/developers: • It is the responsibility of the State Geologist to issue maps depicting zones of required investigation for the hazard of surface fault rupture. These Earthquake Fault Zones (EFZ) are typically 1000 feet wide or greater and are established to encompass identified active faults and an area around those faults where secondary fault traces might lie. • It is the responsibility of the local permitting agency (city or county) to require and approve a geologic report defining and delineating any hazard of surface fault rupture for specified projects within the EFZ. • It is the responsibility of the property owner or developer to cause the property to be investigated by a licensed geologist who will identify the location of any active fault traces so that they may be avoided. For more information on the Alquist-Priolo Earthquake Fault Zoning Act refer to Bryant and Hart (2007).

For our zoning evaluation of the Hollywood Fault within the Hollywood Quadrangle we made use of all available data. This included analysis of vintage aerial photography and topographic maps as well as the compilation and reference to all available geotechnical studies, subsurface data and mapping along the fault zone.

Vintage aerial photographs and maps are a tremendous resource for interpreting the tectonic geomorphology along a fault zone, which is often a key to identifying active fault strands. In areas such as Hollywood this can be a particular challenge because most of the area was already developed, with a grid of streets and structures obscuring the original landform, before the first stereo aerial imagery in 1927. Geologic mapping over the decades had identified some (probably inactive) faults within uplifted bedrock areas, but the more recently active fault traces are mapped as concealed near the interface between the uplifted Santa Monica Mountains and the alluvial fans to the south (Hoots, 1930; Dibblee, 1991). Dolan and others (1997) made great strides in appreciating the tectonic geomorphology of the fault zone. Fault expression and bedrock geology in the Hollywood area. (Modified from Dolan and others, 2000, Fig.2)

Within the western part of the study area we had access to a number of detailed site studies to help constrain age of faulting as well as fault location. These data tended to corroborate geomorphic assessments, but with some minor adjustments needed. In contrast, the eastern part of the study area contained relatively few site-specific studies, but perhaps a more telling geomorphic signature. We were fortunate to have a remarkable topographic survey of the area conducted in 1924 & 1925 that captured the landform with 5-foot contours (USGS, 1926 & 1928). Based on these maps, 1927 and 1928 vintage vertical aerial photography and some historic oblique images from the 1920s and 1930s, we were able to interpret a relatively continuous zone of faults across this eastern area.

Geomorphic features in the Hollywood area with faults that are basis for APEFZ. Local anticlinal fold (green) based on geotechnical studies. Capitol Records Tower indicated by yellow dot. This field trip stop indicated by yellow star. See previous figure for explanation of symbols.

North-south section along Cahuenga Blvd. showing subsurface interpretation of faults based on studies for Los Angeles subway (Crook and Proctor, 1992). Northern and southern strands are indicated on previous figures.

Fault location is only one factor of concern, the other being recency of the latest fault displacement. Several studies in the western part of the map demonstrated that the Hollywood Fault had experienced displacement during the Holocene. Fault influences on drainage and deposition patterns in the eastern part of the map appear to also be in accord with relatively young (although currently unquantified) fault activity. Holocene displacement does not need to be documented on all elements of a fault zone for those traces to be included in a zone of required investigation.

Our evaluation of regional and detailed data indicated that the Hollywood Fault is active and comprises a sometimes complex zone of faults that extends in a generally east-west direction across the Hollywood area (California Geological Survey, 2014). Our findings and conclusions were summarized in a final Fault Evaluation Report after public review of our preliminary recommendations (Hernandez and Treiman, 2014; Hernandez, 2014).

References

Bryant, W.A., and Hart, E.W., 2007, Fault-Rupture Hazard Zones in California, Alquist-Priolo Earthquake Fault Zoning Act with Index to Earthquake Fault Zones Maps: California Geological Survey. Special Publication 42, 42 p. (digital version only, electronic document available at ftp://ftp.consrv.ca.gov/pub/dmg/pubs/sp/Sp42.pdf ). California Geological Survey, 2014, Earthquake Zones of Required Investigation, Hollywood 7.5 Minute Quadrangle: California Geological Survey, 1:24,000. Crook, R., Jr., and Proctor, R.J., 1992, The Santa Monica and Hollywood faults and the southern boundary of the Transverse Ranges Province, in Pipkin, B.W. and Proctor, R.J., eds., Engineering Geology Practice in Southern California: Association of Engineering Geologists, Southern California Section, Special Publication No.4, p.233-246. Dibblee Jr., T.W., 1991, Geologic map of the Hollywood and Burbank (south ½) Quadrangles, Los Angeles County, California: Dibblee Geological Foundation, Map DF-30, 1:24,000. Dolan, J.F., Sieh, K., Rockwell, T.K., Guptill, P. and Miller, G., 1997, Active tectonics, paleoseismology, and seismic hazards of the Hollywood fault, northern Los Angeles basin, California: Geological Society of America Bulletin, v.109, p.1595-1616. Dolan, J.F., Stevens, D., and Rockwell, T.K., 2000, Paleoseismologic evidence for an early to mid-Holocene age of the most recent surface rupture on the Hollywood Fault, Los Angeles, California: Bulletin of the Seismological Society of America, v.90, p.334-344. Hernandez, J.L., 2014, The Hollywood Fault in the Hollywood 7.5’ quadrangle, Los Angeles County, California: California Geological Survey Fault Evaluation Report FER-253, Supplement No.1, November 5, 2014, 35p. Hernandez, J.L., and Treiman, J.A., 2014, The Hollywood Fault in the Hollywood 7.5’ quadrangle, Los Angeles County, California: California Geological Survey, Fault Evaluation Report FER-253, February 14, 2014, 35p. Hoots, H.W., 1930, Geology of the eastern part of the Santa Monica Mountains, Los Angeles County: U.S. Geological Survey Professional Paper 165, p. 83 -134, map scale 1:24,000. USGS, 1926, Burbank, Calif.: United States Geological Survey, 6-minute topographic map series for Los Angeles County, surveyed 1924, 1:24,000. USGS, 1928, Glendale, Calif.: United States Geological Survey, 6-minute topographic map series for Los Angeles County, surveyed 1925, 1:24,000.

(Hernandez and (Hernandez trip stop location. (base map is is map (base location. stop trip

Priolo Earthquake Fault Zone for required investigation investigation required for Zone Fault Earthquake Priolo - traces of the Hollywood Fault Zone across the Hollywood 7.5’ quadrangle 7.5’ Hollywood the across Zone Fault Hollywood the of traces shaded zone surrounding the faults is the Alquist the is faults the zone surrounding shaded

2014). The yellow , ical Survey, 2014). 1925 topographic survey) topographic 1925 - This map shows the interpreted and locally verified active active verified and locally interpreted the map shows This Hernandez 2014; Treiman, (California Geolog Strip map showing features related to fault interpretation in the Hollywood, CA Quadrangle. Yellow star indicates this field this indicates star Yellow Quadrangle. CA Hollywood, the in interpretation fault to related features map showing Strip from 1923

STOP 3: LUNCH AT WATTLES MANSION / RUNYON CANYON PARK From Pantages, drive west on Hollywood Boulevard, and then turn north onto Curson Ave. Our lunch stop takes us back into the Santa Monica Mountains, here a quartz diorite crystalline basement rocks that intruded the Santa Monica Slate exposed to our west.

At this stop we will discuss what is known about the earthquake history of the Hollywood fault and its extensions to the west (Santa Monica fault) and East (Raymond fault). This section of the Hollywood fault was the subject of intensive study during the mid-1990’s as part of the investigation of the subsurface conditions for the LA subway Red Line extension through the Santa Monica Mountains to the San Fernando Valley.

 Map of Hollywood fault and environs (Dolan et al., 1997). Scarps associated with recently active traces of the Hollywood fault shown in reddish-orange. Young alluvial fans (bright colors) are actively being deposited along much of the mountain front (or at least they were being deposited before everything got paved over). Note ~1.5-km-wide left step at west end of the Hollywood fault onto its continuation farther west – the Santa Monica fault. Also note zone of uplifted older alluvium south of the active alluvial front along the Hollywood fault. This uplift is a surface manifestation of uplift above the western, Los Angeles segment of the PHT.

 Geologic map of young features within detailed study area west of downtown Hollywood. Runyon Canyon, Vista Street, and Outpost Drive fans are shown in shades of gray. Narrow, dark gray horizontal swath shows location of Hollywood fault inferred from subsurface data. Medium gray shading shows fault scarps inferred from topography. No scarps are discernible across the recently active parts of the fans. Thick black north-south lines show locations of trenches and borehole transects discussed in text. Secondary strand of Hollywood fault encountered in Fuller Avenue trench is shown by short black line immediately south of borehole OW-34A. Location of Metropolitan Transit Authority subway tunnel excavated as of July 1995 is shown as a dashed line. Triangular facets in northeast corner of figure show possible northeast-trending fault strand. CP-MT—Camino Palmero–Martel Avenue Transect; HA— Hillside Avenue; NLBT—North La Brea transect; WP—Wattles Park; Q shows location of near-surface (<1 m depth) quartz diorite from Crook and Proctor (1992). Topography redrafted from Burbank and Hollywood 1:24 000 6¢ USGS quadrangles (~1926). Contour interval is 1.5 m (5 ft) up to the 500 ft contour, above which the interval is 7.6 m (25 ft).

Dolan et al. (1997) detailed the results of two north-south (i.e., fault-perpendicular) transects of continuously cored hollow-stem boreholes. Twenty-five boreholes comprised the longer, western transect, which extended southward from the mountain front along Camino Palmero and Martell venue for 525 m. Correlation of numerous laterally continuous soils showed that the recently active trace of the Hollywood fault is located just south of the southernmost surface exposures of the quartz diorite that makes up this part of the Santa Monica Mountains.

 Cross section of the northern half of Camino Palmero–Martel Avenue borehole transect. Thick vertical lines denote continuously cored boreholes; thinner lines show sections of B-8 that were not cored. Sub- horizontal black lines denote A and Bt horizons of buried soil horizons of unit C. White zones between these buried soils denote C horizons of buried soils and unaltered sedimentary strata that do not exhibit any soil development. Small triangles and gray lines denote ground-water level in boreholes during 1992. Open circles show locations of two accelerator mass spectrometry–dated detrital charcoal samples discussed in text. Locations of boreholes B-17 and B-16 are projected due east to the line of the cross section. Because of uncertainty of the fault strike, only water-level data from these two boreholes are shown in the figure; stratigraphic data from these holes are not shown in the figure.

Subsequent investigations of the main Hollywood fault crossing by Dolan et al. (2000b) using adjacent large-diameter “bucket-auger” boreholes revealed the details of the alluvial stratigraphy and allowed direct examination of the most recent surface rupture at the site. Bulk-soil radiocarbon ages from both unfaulted and faulted strata demonstrate that the MRE occurred between ~6 ka and 11 ka, with a preferred age range of 7 ka to 9.5 ka.

 Cross section of large-diameter borehole transect discussed in this article. Calibrated 14C ages are all from bulk-soil samples except the two shallowest samples, which are from detrital charcoal fragments. Note the clear fault displacements of the Unit 3/Unit 4 contact in CP-D, CP-C, and CP-F, as well as the lack of displacement of the Unit 2/Unit 3 contact. Also note the 1.3-m-tall, north-side-down step in the Unit 4/Unit 5 contact between CP-G and CP-A. The inferred fault (dashed) shown at this step is drawn at the same dip as that of fault strands observed in CP-D (see text for discussion). The dip of the Unit 4/Unit 5 contacts in CP-C and CP-A was not mapped in detail due to very wet conditions. The average depth of these contacts, however, is accurate to within a few cm. These contacts are shown as dipping the same as the average dip of the Unit 4/Unit 5 contact between CP-H and CP-C. Zigzag lines denote numerous closely spaced fractures not logged in detail due locally unsafe hole conditions. Wavy horizontal lines denote the shallowest groundwater encountered in the boreholes; no groundwater was encountered in boreholes CP-D, CP-G, and CP-I. The irregular dotted line at 5 m depth between CP-G and CP-A shows the area that was hand-excavated from borehole CP-G in order to confirm that the Unit 2/Unit 3 contact was not faulted between the two boreholes. Small-diameter boreholes B-10 and B-12 are from Dolan and others (1997). Only the upper 15 m of these boreholes is shown. They extend to total depths of 73.2 m and 29.4 m, respectively.

STOP 4: AUTRY OVERLOOK ON MULLHOLAND DRIVE

Drive west on Hollywood Boulevard to Laurel Canyon, head north. Near the top turn west onto Mulholland Drive. Bus will cross up and over the Santa Monica Mountains, stopping on Mulholland Drive to overlook the San Fernando Valley (like, totally!) and a view of the mountain ranges (and their faults) to the north.

https://www.flickr.com/photos/cityprojectca/4480269277/

 From this the overlook you can see the broad expanse of the San Fernando Valley. The Simi Hills and Santa Susana Mts form the western and northwestern margins of the valley, respectively, with Oat Mt forming the high point to the northwest (1142 m). The low saddle to the east of Oat Mt (Santa Susana Pass) separates the Santa Susana Mts (west) from the western San Gabriel Mts (east) capped by San Gabriel Pk as the high point (1878 m). Overpasses connecting the Golden State and Antelope Valley freeways (I-5 and Hwy 14) in Santa Susana Pass partially collapsed during both the 1971 and 1994 earthquakes. The Verdugo Mountains (where we started today) are the nearest range to the northeast, in front of the San Gabriel Mts. The north-dipping Sierra Madre Fault System is the longest fault seen from this vantage, including the Santa Susana Section within the Santa Susana Mountains, the San Fernando Section in the middle, and the Sierra Madre section under the San Gabriel Mountains. Splays of this fault system also include the north-dipping Mission Hills fault zone and Northridge Hills (blind) thrust, the subject of our next stop. The 1971 Sylmar earthquake (Mw6.7) ruptured about 15 km of the San Fernando Fault. The 1994 Northridge earthquake (Mw6.7) occurred on a south-dipping blind thrust that projects beneath and terminates against the overriding Northridge Hills fault in the northwest part of the valley.

 Perspective view of terrain from Hollywood Hills to the Garlock Fault, from Q-fault database (http://earthquake.usgs.gov/hazards/qfaults/). Note that the unnamed blind fault responsible for the 1994 Northridge earthquake is not the Northridge Hills Fault! (The 1994 rupture plane terminates against the overriding Northridge Hills fault) SMF=Sierra Madre Fault System.

STOP 5: NORTHRIDGE RECREATION CENTER Drive west along Mulholland Drive, turn onto Coldwater Canyon Ave headed north, and get on the 101 freeway headed west. Then take I-405 N about 8 miles to Nordhoff Street exit, drive west 0.7 miles to Woodley Avenue, turn right. Drive 1.0 mile north on Woodley to Lassen Street, turn left. Drive 2.9 miles west to Reseda Boulevard, turn right. Drive north on Reseda 0.4 miles to Lemarsh Street, turn right. Drive 0.2 miles east on Lemarsh to the Northridge Recreation Center and park in the parking lot to the east of the center. Entrances to public restrooms are located at the west side of the building.

En route to this site we crossed the Northridge Hills fault while heading north on Woodley Avenue (just east of Bull Canyon wash on Figure 1, below). Here, the structure forms an impressive south-facing topographic escarpment. This hill is part of a series of aligned, WNW-trending hills that form the north ridge of the San Fernando Valley, hence the name of the city, the university, and the 1994 earthquake!

It is important to note that the Northridge Hills fault does not reach the earth’s surface. Rather, the Northridge hills are the expression of a fault-propagation fold above the tip of the blind Northridge Hills thrust, located ~1.5 km below this location (see cross-section E-E’ below, from Figure in Tsutsumi and Yeats, 1999).

The purpose of the stop at Lemarsh Street is to visit the Northridge Park site of Baldwin et al. (2000) (Figure 1). Walk southeast from the parking area along a dirt road until you are standing in beneath LA Department of Water and Power’s high-tension power lines. The power lines follow Aliso Canyon where young alluvium fills a late Pleistocene incision into Plio-Pleistocene Saugus Formation (Figure 4). Baldwin et al. (2000) chose to conduct their study at this site because it represents the only place in the Northridge Hills that has not seen significant cultural modification. Here they excavated six test pits, nine boreholes, and one trench (Figure 5). The excavations provided no evidence of surface faulting but do

17 show Holocene growth strata with 13 ± 2 m of vertical relief across a monocline, and a dip-slip rate for the Northridge Hills fault of 1.0 ± 0.7 mm/yr (Figure 6). This study therefore provides direct evidence that the structure deforms Holocene strata, confirming previous interpretations of the faults activity (e.g., Barnhart and Slosson, 1973; Saul, 1979; Dibblee, 1992; Johnson et al., 1996; Hitchock and Kelson, 1996).

 Figure 7 from Tsutsumi and Yeats, 1999. [Note: see Figure 1 of this field guide (Baldwin et al., 2000) for location of cross section line].

 Figures 1, 4, 5, and 6 below are from Baldwin et al. (2000).

STOP 6: CALIFORNIA STATE UNIVERSITY NORTHRIDGE, PARKING LOT F10 Return to the bus/vehicles. Drive 0.2 miles east on Lemarsh Street to Lindley Avenue, turn right. Drive 0.2 miles south and turn left into the northwest entrance to CSUN parking lot F10. Park at base of hill adjacent to dirt path. Walk up the path to reach the top of the hill.

Note that the hill that you climbed has been modified extensively by grading. Resting on top of this hill are ‘relics’ of damage to CSUN’s campus during the 1994 earthquake. The relics are damaged support columns from the Oviatt Library. Rumor has it that the University has stored these ‘ruins’ with plans to use them one day to construct an earthquake memorial on campus. From this vantage point one can see the five mountain ranges that ring the San Fernando Valley (clockwise from north, the San Gabriel, Verdugo, Santa Monica (Autry overlook stop), Simi Hills/Chatsworth and Santa Susana ranges). The saddle at the eastern end of the Santa Monica Mts is Cahuenga Pass, traversed by the Hollywood Freeway (Hwy 101).

In 2005-6 California State University Northridge (CSUN) hired William Lettis and Associates, Inc. to excavate a series of trenches across this parking lot and hill. The trenches total length exceeded 300 m! At the time CSUN was planning to develop the hill with affordable housing units for faculty and staff. However, the economic downturn of 2007-2009 put a halt to these plans. If you look carefully you can still see the patched areas of the parking lot where the trenches were cut. The dirt path that you followed to the hilltop is a scar from one of these trenches. The trenches exposed minor faulting with <10 cm of offset per fault zone, suggesting the site does not experience large ground ruptures during earthquakes (John Baldwin, oral communication, 2006).

The main purpose of this stop is to discuss the tectonic setting of the 1971 Sylmar and 1994 Northridge earthquakes. The epicenters of these earthquakes are 22 km and 4 km to the north of this stop, respectively. The Northridge Hills overlie the north-dipping 1971 Sylmar and south-dipping 1994 Northridge ruptures. Aftershocks define the ruptures in cross section (Figure 7) and map view (Figure 8).

The 1971 earthquake ruptured the San Fernando section of the fault (see figure above). Southwestern splays of the San Fernando section, including the Mission Hills fault zone and the Northridge Hills fault (this stop), did not slip during this earthquake. However, small surface breaks did occur on the north-limb of the Sylmar syncline (Figure 7) interpreted to accommodate bedding-parallel flexural slip (Tsutsumi and Yeats, 1999).

The 1994 earthquake ruptured an un-named south-dipping fault that intersects the Northridge Hills thrust at 5-7 km depth beneath the Sylmar basin (~10 km to the north of this stop). Uplift at this location of ~0.25 m following the 1994 earthquake (Johnson et al., 1996) can be explained by slip on the un-named blind fault triggering slip on the overriding Northridge Hills fault. Opposite-verging thrust systems like this one are common in the region. For example, the blind Northridge/un-named thrust system here is very similar in geometry and style to the opposite-verging Oak Ridge/San Cayetano thrust system 35 km to the northwest of this location, just 7 km deeper.

 Cross section E-E’ from Figure 4, Tsutsumi and Yeats, 1999.

 Figure 8 from Baldwin et al., 2000.

STOP 7: LOMA ALTA PARK Return to Reseda Boulevard and take Hwy 118 E to I-210 E. Take Arroyo Boulevard Exit, and turn north on Windsor Ave, East on Woodbury Road, north on Lincoln Ave and East on Loma Alta Dr. The parking lot is just to north on Sunset Ridge Rd. This part of the tour takes us between the San Gabriel Mountains (to our north) and the Verdugo Mountains (to our south).

The purpose of this stop is to provide a quick overview and description of the Sierra Madre Fault system. We are standing about half way along the Sierra Madre Fault System – recall it has several sections. From east to west, the main fault sections are the Santa Susana Fault (30 km), the San Fernando Fault (20 km), the Sierra Madre Fault (60 km), and the Cucamonga Fault (25 km). The complexity of the fault system leads to a wide range of magnitudes that could rupture on this fault; earthquakes could be moderate (like the M6.7 Sylmar earthquake) or if multiple fault sections are connected, as large as ~M8 (Field et al., 2013).

 Map of Sierra Madre Fault System. BLUE: Santa Susana Fault, 6 (0.5-10) mm/yr; YELLOW: San Fernando Fault, 2 (1-3) mm/yr; RED: Central Sierra Madre Fault, 2 (1-3) mm/yr, PURPLE: Cucamonga Fault 1.5 (1-2) mm/yr. Slip rates used in UCERF3. From Burgette et al., 2014.

 Quaternary geology of the JPL area (Crook et al., 1987), draped over lidar-derived shaded relief. R98 is location of Loma Alta Park. Profiles across mapped traces of the SMF show significant late Quaternary offset (e.g., A-A’). A cross-fan profile (B-B’) illustrates likely miscorrelation of mapped terrace surfaces in earlier work. From Burgette et al. (2014).

Slip rates are variable across this system, and limited in number. For example, the Santa Susana Fault slip rate could be as high as 5.9±3.9 mm/yr (based on offset Plio-Pleistocene units; Huftile and Yeats, 1996) or as low as 2.5 mm/yr based on BEM modeling (Marshall et al., 2013). Lindvall and Rubin (2007) completed cosmogenic dating on the Wilson fan on the San Fernando section; summing offsets on the three strands at that location provides an upper rate of 3 mm/yr. At this location on the Sierra Madre Fault, Crook et al., (1987) estimated slip rates from soil chronosequences of 1.2-3 mm/yr. To the east, Lindvall and Rubin (2007) determined a slip rate of 1.9±0.4 mm/yr across the Cucamonga Fault zone.

There have been a few paleoseismic studies of this system. Here at Loma Alta Park, Rubin et al. (1998) trenched the ~2 m high scarp to our north and see evidence for two paleoearthquakes. They estimate that slip of ~4 m and 6.5 m occurred in the last two earthquakes, which happened since about 18,000 years ago, producing a slip rate of about 0.6 mm/yr (although this is just for one strand of system). Just west of the intersection of the Sierra Madre and Cucamonga Faults, Tucker and Dolan (2001) completed a boring and trench at Horsethief Canyon. They found evidence for one event over 8000 years ago, and determined a slip rate of 0.6-0.9 mm/yr.

 Figures from Rubin et al. (1998)

 1930’s photo of the San Gabriel Mountains with pre-development and relatively tree-free view of uplifted fan surfaces (arrow) above what is now Arroyo Seco.

RETURN TO PASADENA CONVENTION CENTER

References by stop Verdugo Fault: Weber, F. H., Bennett, J.J., Chapman, R.H., Chase, G.W., and Saul, R.B., 1980. Earthquake hazards associated with the Verugo- Eagle Rock and Benedict Canyon Fault Zones, Los Angeles County, Califorinia, USGS Open File Report 81-296.

Griffith Observatory and Wattles Mansion Stop (Dolan) Buwalda, J. P., 1940, Geology of the Raymond Basin: Report to the City of Pasadena Water Dept., 9 plates, 131 p.

Chapman, R. H. and Chase, G. W., 1979, Geophysical investigations of the Santa Monica-Raymond fault zone, Los Angeles County, California, Calif. Div. Mines and Geology, Open-File Report 79-16, p. E-1 to E-30.

Crook, R., and Proctor, R., 1992, The Hollywood and Santa Monica faults and the southern boundary of the Transverse Ranges province, in Pipkin, B., and Proctor, R., eds., Engineering geology practice in southern California: Belmont, California, Star Publishing, p. 233–246.

Dolan, J. F., Sieh, K., Rockwell, T. K., Guptill, P., and Miller, G., 1997, Active tectonics, paleoseismology, and seismic hazards of the Hollywood fault, northern Los Angeles basin, California: Bulletin of the Geological Society of America, v. 109, p. 1595-1616.

Dolan, J. F., Stevens, D., and Rockwell, T. K., 2000, Paleoseismologic evidence for an early to mid-Holocene age of the most recent surface rupture on the Hollywood fault, Los Angeles, California: Bulletin of the Seismological Society of America, v. 90, p. 334-344.

Leon, L. A., Dolan, J. F., Shaw, J. H., and Pratt, T. L., 2009, Evidence for large-magnitude Holocene earthquakes on the Compton blind thrust fault, Los Angeles, California: Jour. Geophys. Res., doi: 10.1029/2008JB006129.

Shaw, J.H., and P. Shearer, (1999). An elusive blind-thrust fault beneath metropolitan Los Angeles, Science, 283, 1516-1518.

Shaw, J. H., Plesch, A., Dolan, J. F., Pratt, T., and Fiore, P., 2002, Puente Hills blind-thrust system, Los Angeles basin, California: Bulletin of the Seismological Society of America, v. 92, p. 2946-2960.

Northridge Stops (Yule)

Baldwin, J., N., Kelson, K.I., and Randolph, C.E., (2000). Late Quaternary fold deformation along the Northridge Hills fault, Northridge, California: Deformation coincident with past Northridge blind-thrust earthquakes and other nearby structures?, Bull. Seism. Soc. Am. 90, 629-642.

Barnhart, J., and J. Slosson (1973). The Northridge Hills and associated faults: a zone of high seismic probability? Assoc. Eng. Geol. Spec. Pubs. 253.

Dibblee, T. W., Jr. (1992). Geologic map of the Oat Mountain and Canoga Park (north 1/2) quadrangles, Los Angeles County, California, Dibblee Geol. Foundation Map DF-36, Santa Barbara, California, scale1:24,000.

Earthquake Engineering Research Institute (1994). Northridge earthquake, January 17, 1994, Preliminary reconnaissance report, J. F. Hall (Editor), California Institute of Technology, Pasadena, California, 96 pp.

Hitchcock, C., and K. Kelson (1999). Growth of late Quaternary folds in southwest Santa Clara Valley, San Francisco Bay area, California:implications of triggered slip for seismic hazard and earthquake recurrence, Geology 27, 5, 391–394.

Johnson, J. A., R. J. Shlemon, and J. E. Slosson (1996). Permanent ground deformation associated with the 17 January 1994 Northridge, California Earthquake; and Geomorphic indicators of potential future zones of deformation, in 1996 Progress Reports from SCEC Scientists, March 1997, Vol II.

Levi, S., and R. S. Yeats (1993). Paleomagnetic constraints on the initiation of uplift on the Santa Susana fault, western Transverse Ranges, California, Tectonics 12, 688–702.

Mori, J., D. J. Wald, and R. L. Wesson (1995). Overlapping fault planes of the 1971 San Fernando and 1994 Northridge, California earthquakes, Geophy. Res. Let. 22, 1033–1036.

Saul, R. (1979). Geology of the southeast quarter of the Oat Mountain Quadrangle, Los Angeles County, California, Div. of Mines and Geology Map Sheet 30, scale 1:12,000, 19 pp.

Tsutsumi, H., and R. S. Yeats (1999). Tectonic setting of the 1971 Sylmar and 1994 Northridge earthquake in the San Fernando Valley, California, Bull. Seism. Soc. Am. 89, 1232–1249.

Wald, D. J., T. H. Heaton, and K. W. Hudnut (1996). The slip history of the 1994 Northridge, California, earthquake determined from strong ground motion, teleseismic, GPS and leveling data, Bull. Seism. Soc. Amer. 86, S49–70.

Wills, C. J., and C. S. Hitchcock (1999). Late Quaternary sedimentationand liquefaction hazard in San Fernando Valley, Los Angeles County,California, Environmental and Engineering Geoscience V., No. 4, 441–458.

Loma Alta Park (Scharer) Burgette, R., Scharer, K., and Hanson, A. (2014). Late Quaternary offset along the Sierra Madre Fault revealed by high-resolution topographic data, GSA Annual Meeting, 323-10.

Lindvall, S.C., & Rubin, C.M., (2007). Slip Rate Studies Along The Sierra Madre-Cucamonga Fault System Using Geomorphic and Cosmogenic Surface Exposure Age Constraints. NEHRP Final Technical Report, award 03HQGR0084, 13 p.

Marshall, S. T., Funning, G. J., & Owen, S. E. (2013). Fault slip rates and interseismic deformation in the western Transverse Ranges, California. Journal of Geophysical Research: Solid Earth, 118(8), 4511-4534.

Rubin, C. M., Lindvall, S. C., & Rockwell, T. K. (1998). Evidence for large earthquakes in metropolitan Los Angeles. Science, 281(5375), 398-402.

Tucker, A. Z., & Dolan, J. F. (2001). Paleoseismologic evidence for a> 8 ka age of the most recent surface rupture on the eastern Sierra Madre fault, northern Los Angeles metropolitan region, California. Bulletin of the Seismological Society of America, 91(2), 232-249.

Yerkes, R. F., and Campbell, R. H., comp (2005). Preliminary Geologic Map of the Los Angeles 30” x 60’ Quadrangle, Southern California, USGS Open File Report 2005-0019, http://pubs.usgs.gov/of/2005/1019/.

Bulletin of the Seismological Society of America, Vol. 92, No. 8, pp. 2946–2960, December 2002

Puente Hills Blind-Thrust System, Los Angeles, California by John H. Shaw, Andreas Plesch, James F. Dolan, Thomas L. Pratt, and Patricia Fiore

Abstract We describe the three-dimensional geometry and Quaternary slip history of the Puente Hills blind-thrust system (PHT) using seismic reﬂection proﬁles, petroleum well data, and precisely located seismicity. The PHT generated the 1987

Whittier Narrows (moment magnitude [Mw] 6.0) earthquake and extends for more than 40 km along strike beneath the northern Los Angeles basin. The PHT comprises three, north-dipping ramp segments that are overlain by contractional fault-related folds. Based on an analysis of these folds, we produce Quaternary slip proﬁles along each ramp segment. The fault geometry and slip patterns indicate that segments of the PHT are related by soft-linkage boundaries, where the fault ramps are en echelon and displacements are gradually transferred from one segment to the next. Average Quaternary slip rates on the ramp segments range from 0.44 to 1.7 mm/yr, with preferred rates between 0.62 and 1.28 mm/yr. Using empirical relations among rupture area, magnitude, and coseismic displacement, we estimate the magnitude and

frequency of single (Mw 6.5–6.6) and multisegment (Mw 7.1) rupture scenarios for the PHT.

Introduction

The Los Angeles basin lies along the southern Califor- of the PHT with seismic reflection profiles and petroleum nia coast at the junction of the Transverse and Peninsular well data. Using precisely relocated seismicity, these authors Ranges. The basin is currently being deformed by several proposed that the Santa Fe Springs segment of the PHT gen- blind-thrust and strike-slip faults that have generated his- erated the 1987 (Mw 6) Whittier Narrows earthquake. In this toric, moderate-size earthquakes (Hauksson, 1990; Dolan et article, we define the size and three-dimensional geometry al., 1995). Geodetic studies suggest that the northern basin of the PHT using additional seismic reflection data. We de- is shortening at a rate of 4.4–5 mm/yr in a north–south to scribe the systematic behavior of the segments that compose northeast–southwest direction (Walls et al., 1998; Argus et the PHT and consider the geometric and kinematic relations al., 1999; Bawden et al., 2001). Part of this shortening is of the PHT to other blind-thrust and strike-slip systems. Fi- accommodated on recognized fault systems, and two com- nally, we map the distribution of slip on the PHT and discuss peting models have been proposed to explain which faults the implications of this study for earthquake hazard assess- account for the remainder. Walls et al. (1998) suggested that ment in metropolitan Los Angeles. the remaining shortening is accommodated by higher rates of slip on conjugate strike-slip systems in the northern basin. Motion on these faults produces north–south shortening by Puente Hills Blind Thrust lateral crustal extrusion. In contrast, Argus et al. (1999) and Displacements on the PHT contribute to the growth of Bawden et al. (2001) proposed that the shortening is accom- a series of en echelon anticlines in the northern Los Angeles modated primarily by activity on thrust and reverse faults. basin that involve Miocene through Quaternary strata. The In this article, we document an active blind-thrust system in easternmost fold forms the Coyote Hills and provides struc- the northern Los Angeles basin, termed the Puente Hills tural closure for the east and west Coyote Hills oil fields blind thrust (PHT), that accommodates a component of the (Yerkes, 1972). The central fold has only subtle surface ex- unresolved basin shortening and poses substantial earth- pression and provides structural closure for the Santa Fe quake hazards to metropolitan Los Angeles. Springs oil field (Fig. 1). The westernmost fold is contained The PHT extends for more than 40 km along strike in within a broad, south-dipping monocline that forms the the northern Los Angeles basin from downtown Los Angeles northern boundary of the Los Angeles basin (Davis et al., east to Brea in northern Orange County (Fig. 1). The fault 1989; Shaw and Suppe, 1996; Schneider et al., 1996). All consists of at least three distinct geometric segments, termed of these folds are bounded on their southern margins by Los Angeles, Santa Fe Springs, and Coyote Hills, from west narrow forelimbs (Fig. 2). These forelimbs, or kink bands, to east. Shaw and Shearer (1999) defined the central portion consist of concordantly folded, south-dipping Pliocene

2946 Tom Pratt (USGS), James Dolan (USC) and John Shaw (Harvard) collecting high-resolution seismic re ection data above the central, Santa Fe Springs segment of the Puente Hills thrust (see Pratt et al., 2002 (BSSA); Shaw et al. 2002 (BSSA); and Dolan et al., 2004 (Science).

! JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 112, B03S03, doi:10.1029/2006JB004461, 2007 Click Here JOURNAL OF GEOPHYSICALfor RESEARCH, VOL. 112, B03S03, doi:10.1029/2006JB004461, 2007 Click Here Full for Article B03S03 LEON ET AL.: PUENTE HILLS THRUST EARTHQUAKE B03S03 Full B03S03 LEON ET AL.: PUENTE HILLS THRUST EARTHQUAKE B03S03 Article Earthquake-by-earthquake fold growth above the Puente Hills blind Earthquake-by-earthquake fold growth above the Puente Hills blind thrust fault, Los Angeles, California:thrust fault, Implications Los Angeles, for fold California: Implications for fold kinematics and seismic hazardkinematics and seismic hazard Lorraine A. Leon,1 Shari A. Christofferson,Lorraine1 James A. Leon, F. Dolan,1 Shari1 A. Christofferson,1 James F. Dolan,1 John H. Shaw,2 and Thomas L. PrattJohn3 H. Shaw,2 and Thomas L. Pratt3 Received 22 April 2006; revised 7 November 2006; accepted 27 February 2007; published 29 March 2007. Received 22 April 2006; revised 7 November 2006; accepted 27 February 2007; published 29 March 2007. [1] Boreholes and high-resolution seismic reflection data collected across the forelimb growth triangle above the central segment[1] Boreholes of the Puente and Hills high-resolution thrust fault (PHT) seismic beneath reflection data collected across the forelimb Los Angeles, California, provide agrowth detailed triangle record of above incremental the central fold growth segment during of the Puente Hills thrust fault (PHT) beneath large earthquakes on this major blindLos thrust Angeles, fault. These California, data document provide fold a detailed growth record of incremental fold growth during within a discrete kink band that narrows upward from 460 m at the base of the large earthquakes� on this major blind thrust fault. These data document fold growth Quaternary section (200–250 m depth)within to <150a discrete m at 2.5 kink m depth,band thatwith mostnarrows growth upward from 460 m at the base of the during the most recent folding event occurring within a zone only 60 m wide. These � observations, coupled with evidenceQuaternary from petroleum section industry (200–250 seismic� reflection m depth) data, to <150 m at 2.5 m depth, with most growth demonstrate that most (>82% at 250during m depth) the folding most and recent uplift folding occur within event discrete occurring kink within a zone only 60 m wide. These bands, thereby enabling us to developobservations, a paleoseismic coupled history of with the underlying evidence blind from thrust petroleum industry seismic� reflection data, fault. The borehole data reveal thatdemonstrate the youngest part that of most the growth (>82% triangle at 250 in m the depth) folding and uplift occur within discrete kink uppermost 20 m comprises three stratigraphicallybands, thereby discrete enabling growth us intervals to develop marked a paleoseismic by history of the underlying blind thrust southward thickening sedimentary stratafault. that The are borehole separated by data intervals reveal in that which the sediments youngest part of the growth triangle in the do not change thickness across the site. We interpret the intervals of growth as occurring after the formation of now-buried paleofolduppermost scarps 20 during m comprises three large three PHT earthquakesstratigraphically in discrete growth intervals marked by the past 8 kyr. The intervening intervalssouthward of no growth thickening recordsedimentary periods of structural strata that are separated by intervals in which sediments quiescence and deposition at the regional,do not near-horizontal change thickness stream gradientacross the at the site. study We site. interpret the intervals of growth as occurring Minimum uplift in each of the scarp-formingafter the events, formation which of occurred now-buried at 0.2–2.2 paleofold ka (event scarps during three large PHT earthquakes in Y), 3.0–6.3 ka (event X), and 6.6–8.1the ka past (event 8 kyr. W), The ranged intervening from 1.1 intervalsto 1.6 m, of no growth record periods of structural indicating minimum thrust displacements of 2.5 to 4.5 m. Such large� displacements� are quiescence� and deposition at the regional, near-horizontal stream gradient at the study site. consistent with the occurrence of large-magnitude earthquakes (Mw > 7). Cumulative, minimum uplift in the past threeMinimum events was uplift3.3 to in 4.7 each m, suggesting of the scarp-forming cumulative events, which occurred at 0.2–2.2 ka (event thrust displacement of 7 to 10.5Y), m. 3.0–6.3 These values ka (event yield a X), minimum and 6.6–8.1 Holocene ka slip (event W), ranged from 1.1 to 1.6 m, rate for the PHT of 0.9� to 1.6 mm/yr.indicating The minimum borehole and thrust seismic displacements reflection data of 2.5 to 4.5 m. Such large� displacements� are � � demonstrate that dip within the kinkconsistent band is with acquired the incrementally, occurrence of such large-magnitude that older earthquakes (Mw > 7). Cumulative, strata that have been deformed byminimum more earthquakes uplift in dip the more past steeply three than events younger was 3.3 to 4.7 m, suggesting cumulative strata. Specifically, strata dip 0.4� thrustat 4 m displacement depth, 0.7� at 20 of m depth,7 to 810.5� at 90 m. m, These 16� values yield a minimum Holocene slip at 110 m, and 17� at 200 m. Moreover, structural restorations of the borehole data rate for the PHT of 0.9� to 1.6 mm/yr. The borehole and seismic reflection data show that the locus of active folding (the anticlinal active� axial surface) does not extend to the surface in exactly thedemonstrate same location that from dip earthquake within the to earthquake. kink band is acquired incrementally, such that older Rather, that the axial surfaces migratestrata from that earthquake have been to earthquake, deformed reflecting by more a earthquakes dip more steeply than younger component of fold growth by kinkstrata. band migration. Specifically, The incremental strata dip acquisition 0.4� at 4 of m bed depth, 0.7� at 20 m depth, 8� at 90 m, 16� dip in the growth triangle may reflectat 110 some m, combination and 17� at of 200 fold m.growth Moreover, by limb structural restorations of the borehole data rotation in addition to kink band migration, possibly through a component of trishear or shear fault bend folding. Alternatively,show that the thecomponent locus of limb active rotation folding may (the anticlinal active axial surface) does not result from curved hinge fault bendextend folding, to and/or the surfacethe mechanical in exactly response the of same loosely location from earthquake to earthquake. consolidated granular sediments inRather, the shallow that subsurface the axial to surfaces folding at migrate depth. from earthquake to earthquake, reflecting a component of fold growth by kink band migration. The incremental acquisition of bed dip in the growth triangle may reflect some combination of fold growth by limb 1Department of Earth Sciences, University of Southern California, Los Angeles, California, USA. rotation in addition to kink band migration, possibly through a component of trishear 2Department of Earth and Planetary Sciences, Harvard University, Cambridge, Massachusetts, USA. or shear fault bend folding. Alternatively, the component of limb rotation may 3U. S. Geological Survey, Seattle, Washington,result USA. from curved hinge fault bend folding, and/or the mechanical response of loosely consolidated granular sediments in the shallow subsurface to folding at depth. Copyright 2007 by the American Geophysical Union. 0148-0227/07/2006JB004461$09.00 Figure 5. Borehole resultsFigure from 5. theBorehole Carfax Avenue results from transect, the Carfax Gardenland-Greenhurst Avenue transect, transect,Gardenland-Greenhurst and SCE transect, and SCE 1Department ofB03S03 Earth Sciences, University of Southern California, Los 1 of 18 transectAngeles, California, showing USA. majortransect stratigraphic showing units major (8X stratigraphic vertical exaggeration). units (8X vertical Thin vertical exaggeration). lines denote Thin vertical lines denote 2Department of Earth and Planetaryboreholes. Sciences, Harvard Numbers University, show key stratigraphic units discussed in the text. Thin red lines mark the tops of boreholes.Cambridge, Massachusetts, Numbers USA. show key stratigraphic units discussed in the text. Thin red lines mark the tops of major3U. S.sand- Geological and Survey, gravel-filled Seattle, Washington,major units. sand- USA. Double-headed and gravel-filled red units. vertical Double-headed arrows along red the vertical left side arrows of the along figure the left side of the figure show the stratigraphic range within which discrete uplift events occurred (see text for discussion). Solid showCopyright the 2007 stratigraphic by the American Geophysical range Union.within which discrete uplift events occurred (see text for discussion). Solid green0148-0227/07/2006JB004461$09.00 lines between redgreen arrows lines on between left side red of figure arrows show on left stratigraphic side of figure intervals show that stratigraphic do not change intervals that do not change thickness across the transect, reflecting periods of structural quiescence. Proposed correlations between thickness across the transect, reflecting periodsB03S03 of structural quiescence. Proposed correlations1 of 18 between Carfax boreholes and aCarfax water well boreholes (1589S) and located a water 175 well m (1589S) north of located the transect 175 m are north shown of bythe dashed transect are shown by dashed stratigraphic contacts alongstratigraphic the right side contacts of the along figure. the Contacts right side at of 12, the 17, figure. and 20 Contacts m depth at are 12, correlative 17, and 20 m depth are correlative between borehole 22 and the water well. B03S03 between borehole 22LEON and ETthe AL.: water PUENTE well. HILLS THRUST EARTHQUAKE B03S03 16 boreholes drilled to1611 boreholes to 50 m drilled depths to (Figure11 to 5). 50 mwere depths used (Figure for sampling. 5). were The used newly for acquired sampling. Gardenland The newly acquired Gardenland � Initially interpreted by InitiallyDolan� et interpreted al. [2003], by thisDolan borehole et al. [2003],borehole this transect, borehole whichborehole consisted transect, of seven, which continuously consisted of seven, continuously transect was excavated alongtransect the was same excavated alignment along as the the high- same alignmentcored hollow as the stem high- boreholes,cored hollow was drilled stem parallelboreholes, to wasand drilled parallel to and resolution hammer seismicresolution reflection hammer profile seismic (Figure reflection 3c). profile100 m west(Figure of, the3c). Carfax100 Avenue m west transect of, the (Figure Carfax Avenue5). The transect (Figure 5). The � These boreholes thus provideThese boreholes direct ground thus truth provide on direct the SCE� ground transect, truth which on the wasSCE much transect, shorter which (80 m) was than much the other shorter (80 m) than the other youngest reflectors imagedyoungest by reflectorsseismic reflection imaged by data seismictwo transects, reflection was data collectedtwo transects,50 m east was of and collected parallel50 to them east of and parallel to the � (Figure 6). Seven of these(Figure boreholes 6). Seven (18 through of these 24) boreholes were (18northern through part 24) of were the Carfaxnorthern� transect. part of The the SCE Carfax transect transect. The SCE transect continuously sampledcontinuously hollow stem sampled boreholes, hollow whereas stemconsisted boreholes, of whereas seven, large-diameterconsisted of (70 seven, cm) large-diameter bucket auger (70 cm) bucket auger eight of the boreholes (10eight through of the 17) boreholes were large-diameter (10 through 17)holes were excavated large-diameter to 12 m inholes depth excavated (Figure 5).to 12 These m in results depth (Figureare 5). These results are 1 (70 cm) bucket auger holes.(70 The cm) bucket bucket augers auger holes. were sampled The bucketrecorded augers were in the sampled auxiliaryrecorded material1 in. the auxiliary material . downhole directly bydownhole a geologist directly to the by depth a geologist of the to the depth of the 1 shallowest aquifer. Belowshallowest this depth, aquifer. one foot Below bucket this drives depth, one1 footAuxiliary bucket materials drives are availableAuxiliary in materials the HTML. are doi:10.1029/ available in the HTML. doi:10.1029/ 2006JB004461. 2006JB004461.

7 of 18 7 of 18

Figure 5. Borehole results from the Carfax Avenue transect, Gardenland-Greenhurst transect, and SCE transect showing major stratigraphic units (8X vertical exaggeration). Thin vertical lines denote boreholes. Numbers show key stratigraphic units discussed in the text. Thin red lines mark the tops of major sand- and gravel-filled units. Double-headed red vertical arrows along the left side of the figure show the stratigraphic range within which discrete uplift events occurred (see text for discussion). Solid green lines between red arrows on left side of figure show stratigraphic intervals that do not change thickness across the transect, reflecting periods of structural quiescence. Proposed correlations between Carfax boreholes and a water well (1589S) located 175 m north of the transect are shown by dashed stratigraphic contacts along the right side of the figure. Contacts at 12, 17, and 20 m depth are correlative between borehole 22 and the water well.

16 boreholes drilled to 11 to 50 m depths (Figure 5). were used for sampling. The newly acquired Gardenland Initially interpreted by Dolan� et al. [2003], this borehole borehole transect, which consisted of seven, continuously transect was excavated along the same alignment as the high- cored hollow stem boreholes, was drilled parallel to and resolution hammer seismic reflection profile (Figure 3c). 100 m west of, the Carfax Avenue transect (Figure 5). The These boreholes thus provide direct ground truth on the SCE� transect, which was much shorter (80 m) than the other youngest reflectors imaged by seismic reflection data two transects, was collected 50 m east of and parallel to the (Figure 6). Seven of these boreholes (18 through 24) were northern part of the Carfax� transect. The SCE transect continuously sampled hollow stem boreholes, whereas consisted of seven, large-diameter (70 cm) bucket auger eight of the boreholes (10 through 17) were large-diameter holes excavated to 12 m in depth (Figure 5). These results are (70 cm) bucket auger holes. The bucket augers were sampled recorded in the auxiliary material1. downhole directly by a geologist to the depth of the shallowest aquifer. Below this depth, one foot bucket drives 1Auxiliary materials are available in the HTML. doi:10.1029/ 2006JB004461.

7 of 18 R EPORTS

from additional tension in the wires. With a However, polishing of GaAs/AlGaAs superlattices is 12. A. N. Cleland, M. L. Roukes, J. Appl. Phys. 92, 2758 reasonable reduction of resonator lengths to 0.25 well established for laser devices and may be useful (2002). here. 13. In a strong magnetic field, current flowing through the �m, it should be possible to fabricate resonators 6. The brightness/contrast of the SEM images was wires creates a Lorentz force orthogonal to the magnetic field and current directions, in turn creating an with �res � 2.5 GHz. adjusted, but no other image processing was used. As a final demonstration of the flexibility of 7. Y. Huang, X. F. Duan, Q. Q. Wei, C. M. Lieber, Science oscillatory force on the wire with the frequency of the 291, 630 (2001). applied ac current. At resonance, the wire oscillates the SNAP process, the process was repeated to though the magnetic field with appreciable amplitude, 8. M. R. Diehl et al., Angew. Chem. Int. Ed. 41, 353 fabricate ultradense crossbar circuit elements. As producing an induction current that can be measured (2001). with a network analyzer. To enhance sensitivity, we discussed above, one intrinsic advantage of the 9. Typical Si and poly-Si thin films were 30 to 50 nm used a current-sensitive bridge circuit to reduce back- SNAP process is that it is a direct transfer of thick and were etched using Cl2 RIE at low power (80 ground and reflected intensity. prefabricated nanowires, enabling many patterns W ) and pressure (5 mT ) to carefully control the etch 14. P. Mohanty et al., Phys. Rev. B 66, 085416 (2002). rate. 15. J. R. Heath, P. J. Kuekes, G. Snider, R. S. Williams, to be deposited on top of one another. A second 10. Contacts were made by using EBL to define �100- Science 280, 1716 (1998). advantage of the SNAP process is related to the nm wires placed on either side of the etched portion 16. Yi Luo et al., ChemPhysChem 2002, 519 (2002). extremely high aspect ratio of the nanowire ar- of the wire array. Accurate placement of the contact 17. A. Bachtold, P. Hadley, T. Nakanishi, C. Dekker, rays. Because the arrays are so long (mm wires was achieved with the use of predeposited Science 294, 1317 (2001). alignment markers. Wires with narrow pitch (�80 18. Y. Huang et al., Science 294, 1313 (2001). lengths), the only critical alignment between a nm) could generally not be individually addressed 19. P. J. Kuekes, R. S. Williams, U.S. Patent 6,256,767 top and bottom pattern is the angle between the this way because of the resolution limits of EBL. (2001). two patterns. In other words, no particular spatial Contact of a single wire was achieved by depositing 20. Supported by the Office of Naval Research and the alignment is needed even to fabricate crossbar electrodes onto all the wires in parallel, then using a Defense Advanced Research Projects Agency. We focused ion beam to cut all but one wire between the 11 thank M. Roukes and his group for teaching us how to structures at densities in excess of 10 junctions electrodes. perform high-frequency nanomechanical resonator cm�2. Figure 4C is an electron micrograph of 11. The metal nanowires (�15 nm wide) exhibited con- measurements. several crossbar circuits fabricated by repeating sistent bulk metallic conductivity to 4 K; smaller R EPORTS wires could not be measured individually. The con- 27 December 2002; accepted 4 March 2003 the SNAP process at 90° to each other. An from additional tension in the wires. With a However, polishing of GaAs/AlGaAsductivity superlattices of the SOI nanowires is 12. fluctuated A. N. Cleland, much M. more L. Roukes, J.Published Appl. Phys. online92, 2758 13 March 2003; obvious limitation of the SNAP process is that it widely, depending on the processing conditions and 10.1126/science.1081940 reasonable reduction of resonator lengths to 0.25 well established for laser devices and may be useful (2002). can only produce crossbar circuitshere. (or other rel- nanowire crystallographic orientation.13. In a strong magnetic field, currentInclude flowing this through information the when citing this paper. m, it should be possible to fabricate resonators � atively simple structures). However,6. The brightness/contrast the crossbar of the SEM images was wires creates a Lorentz force orthogonal to the mag- with � � 2.5 GHz. adjusted, but no other image processing was used. netic field and current directions, in turn creating an res has emerged as the circuit of choice for a number As a final demonstration of the flexibility of 7. Y. Huang, X. F. Duan, Q. Q. Wei, C. M. Lieber, Science oscillatory force on the wire with the frequency of the of nanoelectronic applications (29115),, 630 such (2001). as mo- applied ac current. At resonance, the wire oscillates the SNAP process, the process was repeated to though the magnetic field with appreciable amplitude, on October 13, 2008 lecular switch–based random8. M. access R. Diehl memoryet al., Angew. Chem. Int.Recognition Ed. 41, 353 of Paleoearthquakes fabricate ultradense crossbar circuit elements. As producing an induction current that can be measured (16) or nanowire-based logic circuits(2001). (17, 18). with a network analyzer. To enhance sensitivity, we discussed above, one intrinsic advantage of the Many applications are possible9. Typical with Si theandcom- poly-Si thin films wereon 30 to 50 the nm Puenteused a current-sensitive Hills bridge circuit Blind to reduce back- Thrust SNAP process is that it is a direct transfer of thick and were etched using Cl2 RIE at low power (80 ground and reflected intensity. bination of nanowire materials, diameters, pitch- prefabricated nanowires, enabling many patterns W ) and pressure (5 mT ) to carefully control the etch 14. P. Mohanty et al., Phys. Rev. B 66, 085416 (2002). es, and aspect ratios that therate. SNAP approach 15. J. R. Heath, P. J. Kuekes, G. Snider, R. S. Williams, to be deposited on top of one another. A second Fault, California yields. Possibilities include10. one-dimensional Contacts were made su- by using EBL to define �100- Science 280, 1716 (1998). advantage of the SNAP process is related to the nm wires placed on either side of the etched portion perconductors, high-density semiconductor 16. Yi Luo1 et al., ChemPhysChem 2002, 519 (2002).1 2 extremely high aspect ratio of the nanowire ar- of the wire array. Accurate placement ofJames the contact F. Dolan,17. A. Bachtold,* Shari P. A. Hadley, Christofferson, T. Nakanishi, C. Dekker,John H. Shaw rays. Because the arraysnanowire are so sensor long arrays, (mm gigahertzwires nanomechani- was achieved with the use of predeposited Science 294, 1317 (2001). cal resonators, and high-densityalignment molecular markers. elec- Wires with narrowBorehole pitch data (�80 from18. young Y. Huang sedimentset al., Science folded294, 1313 above (2001). the Puente Hills blind thrust lengths), the only critical alignment between a www.sciencemag.org tronics circuits. However, onenm) critical could issue generally that not be individuallyfault beneath addressed Los19. Angeles P. J. Kuekes, reveal R. S. that Williams, the folding U.S. Patent extends 6,256,767 to the surface as a top and bottom pattern is the angle between the this way because of the resolution limits of EBL. � (2001). two patterns. In other words,must no be particular faced if spatial this approachContact is to of be a single fully wire was achieveddiscrete by depositing zone ( 20.145 Supported meters by wide). the Office Buried of Naval fold Research scarps and the within an upward- alignment is needed evenexploited to fabricate is the crossbar task of individuallyelectrodes addressing onto all the wires in parallel,narrowing then using zone a of deformation,Defense Advanced which Research extends Projects from Agency. the Weupward termination focused ion beam to cut all but one wire between the each nanowire11 within a small pitch array. The of the thrust ramp atthank 3 kilometers M. Roukes and depth his group to for the teaching surface, us how document to the occur- structures at densities in excess of 10 junctions electrodes. perform high-frequency nanomechanical resonator cm�2. Figure 4C is an electronSNAP technique micrograph can of generate11. The wiring metal networks nanowires (�15 nm wide)rence exhibited of at con- least fourmeasurements. large (moment-magnitude 7.2 to 7.5) earthquakes on this several crossbar circuitsthat fabricated are well by repeating beyond the resolutionsistent bulk of metallic even conductivityfault to 4 during K; smaller the past 11,000 years. Future events of this type pose a seismic EBL. Thus, this approach bringswires into could focus not bethe measured individually.hazard Theto metropolitan con- 27 December Los Angeles. 2002; accepted Moreover, 4 March the 2003 methods developed in this the SNAP process at 90° to each other. An ductivity of the SOI nanowires fluctuated much more Published online 13 March 2003;

obvious limitation of theproblem SNAP process of interfacing is that it the micrometerwidely, depending or sub- on the processingstudy conditions can be and used to refine10.1126/science.1081940 seismic hazard assessments of blind thrusts in other Downloaded from can only produce crossbarmicrometer circuits (or world otherof rel- lithographynanowire and crystallographicthe mac- orientation.metropolitan regions.Include this information when citing this paper. atively simple structures).romolecular-type However, the crossbar densities achievable with has emerged as the circuitthis of choice technique. for a number This is one of the outstanding During the past 30 years, paleoseismology (the and shortening in the near-surface is accommo- of nanoelectronic applicationsproblems (15), in such nanotechnology, as mo- but potential so- study of ancient earthquakes) has provided a dated by folding rather than by fault slip (3–6). on October 13, 2008 lecular switch–based randomlutions, access such memory as binary tree demultiplexers,Recognitionwealth of information of Paleoearthquakes on the locations, magni- The absence of paleoearthquake age and dis- (16) or nanowire-based logichave circuits been proposed (17, 18). (19). Building such de- tudes, and dates of earthquakes (1, 2). These data placement data from blind thrust faults creates a Many applications aremultiplexers, possible with the with com- less stringenton (�50 nm) theare Puente critical for probabilistic Hills seismic Blind hazard as- Thrustmajor gap both in our understanding of the bination of nanowire materials,requirements diameters, in pitch-terms of the pitch and align- sessments, which dictate modern hazard zoning, collective behavior of regional fault systems es, and aspect ratios thatment the of SNAP the electrical approach connections (an essential emergencyFault, response, California and risk mitigation strate- and in our ability to construct accurate seismic yields. Possibilities includerequirement), one-dimensional will be su- a major challenge. gies. Previous paleoseismological studies, how- hazard models. Here we document the ages perconductors, high-density semiconductor James F. Dolan,ever, have1* Shari largely A. neglected Christofferson, an entire class1 John of H.and Shaw displacements2 of individual paleoearth- nanowire sensor arrays, gigahertz nanomechani- faults—the so-called blind thrust faults. These quakes on a major, active blind thrust fault. References and Notes cal resonators, and high-density1. C. Vieu molecularet al., Appl. elec- Surf. Sci. 164, 111Borehole (2000). data fromfaults young do not sediments extend all folded the way above to the the surface, Puente Hills blindOur studythrust focused on the Puente Hills thrust www.sciencemag.org tronics circuits. However,2. one Y. Xia criticalet al., issueScience that273, 347 (1996).fault beneath Los Angeles reveal that the folding extends to the(PHT), surface a major as a blind fault beneath metropolitan must be faced if this approach3. S. Y. Chou, is to P. be R. Krauss, fully P. J. Renstrom,discreteScience zone272, (�145 meters wide). Buried fold scarps withinLos an Angeles, upward- California (7). Extensive petro- 1 exploited is the task of individually85 (1996). addressing narrowing zone ofDepartment deformation, of Earth which Sciences, extends University from of the South- upwardleum-industry termination seismic reflection and well data, 4. We have reproduced this technique on more than a ern California, Los Angeles, CA 90089–0740, USA. each nanowire within a small pitch array. The 2 relocated earthquakes, and geomorphologic ob- dozen commercially purchased andof grown-in-house the thrust rampDepartment at 3 kilometers of Earth and depth Planetary to the Sciences, surface, Harvard document the occur-R EPORTS SNAP technique can generatesuperlattices, wiring networks and we have found thatrence there of is appre- at least fourUniversity, large 20 (moment-magnitude Oxford Street, Cambridge, 7.2 toMA 7.5) 02138, earthquakesservations on demonstratethis that the PHT, which dips that are well beyond the resolutionciable chemical of variability even from waferfault to wafer, during with theUSA. past 11,000 years. Future events of this type posegently a northeastwardseismic and terminates upward at a near-horizontal paleostream gradient.a This corresponding infer- variabilityThe in etch large chemistry. displacements associated with been larger than any earthquakes that have ence is supportedEBL. Thus, by the this laterally approach constant5. brings Currently into thick- the focus superlattice theEvents is not V being throughhazard reused because to Y metropolitan suggest*To whom that Los correspondencethe Angeles. PHT Moreover, shouldoccurred be the addressed. on methods Los E- Angeles developeddepth–region of in� this3 km, faults extends (ex- for 40 to 50 km from ness of Unitsproblem 46 and of interfacing 47, which the indicates micrometerof the small that oramount sub-typically necessary ruptures forstudy each can imprint. in be large usedmail: earthquakes to refine [email protected] seismic (Mw hazardcluding assessments the Sanof blind Andreas thrustsnorthern infault) Orangeother during County, the beneathDownloaded from downtown Los micrometer world of lithography and the mac- metropolitan regions. these sediments were deposited near-horizontal- �7). These PHT earthquakes would have 200-year-long historic period (20). The rec- romolecular-type densities achievable with www.sciencemag.org SCIENCE VOL 300 4 APRIL 2003 115 ly. Units 46this and technique. 47 record This a period is one of of structural the outstanding During the past 30 years, paleoseismology (the and shortening in the near-surface is accommo- quiescence,problems with no indiscernible nanotechnology, uplift occurring but potential so- study of ancient earthquakes) has provided a dated by folding rather than by fault slip (3–6). between 7.8lutions, thousand such years as ago binary (ka) tree and demultiplexers,�9.5 to wealth of information on the locations, magni- The absence of paleoearthquake age and dis- 10.0 ka (14have). In been contrast, proposed the overlying (19). Building interval such de- tudes, and dates of earthquakes (1, 2). These data placement data from blind thrust faults creates a comprisingmultiplexers, the Unit 40 with sand less and stringent the Unit (� 4550 nm) are critical for probabilistic seismic hazard as- major gap both in our understanding of the overbank mudsrequirements thickens in terms southward of theby pitch 2 and m align- sessments, which dictate modern hazard zoning, collective behavior of regional fault systems across the zonement of of active the electrical folding. connections The absence (an of essential emergency response, and risk mitigation strate- and in our ability to construct accurate seismic requirement), will be a major challenge. gies. Previous paleoseismological studies, how- hazard models. Here we document the ages growth in the Units 30 to 35 section supports our ever, have largely neglected an entire class of and displacements of individual paleoearth- inference that the top of the Unit 40 channel sand faults—the so-called blind thrust faults. These quakes on a major, active blind thrust fault. was depositedReferences at the near-horizontal and Notes paleostream 1. C. Vieu et al., Appl. Surf. Sci. 164, 111 (2000). faults do not extend all the way to the surface, Our study focused on the Puente Hills thrust gradient. Thus,2. Y. theXia et sedimentary al., Science 273, growth 347 (1996). within (PHT), a major blind fault beneath metropolitan the Units 403. to S. Y. 45 Chou, section P. R. Krauss, records P. J. Renstrom, onlap ofScience a 272, Los Angeles, California (7). Extensive petro- 85 (1996). 1Department of Earth Sciences, University of South- leum-industry seismic reflection and well data, now-buried southfacing4. We have reproduced fold scarp this that technique developed on more than a ern California, Los Angeles, CA 90089–0740, USA. relocated earthquakes, and geomorphologic ob- sometime afterdozen the deposition commercially of purchased Unit 46 and at 7.8grown-in-house to 2Department of Earth and Planetary Sciences, Harvard 8.2 ka and beforesuperlattices, the completion and we have of found deposition that there of is appre- University, 20 Oxford Street, Cambridge, MA 02138, servations demonstrate that the PHT, which dips ciable chemical variability from wafer to wafer, with USA. gently northeastward and terminates upward at a Unit 40 before 6.6a corresponding to 6.9 ka(the variability age ofin etch the chemistry.middle depth of �3 km, extends for 40 to 50 km from of overlying5. Unit Currently 35). the We superlattice refer is to not this being strati- reused because *To whom correspondence should be addressed. E- of the small amount necessary for each imprint. mail: [email protected] northern Orange County, beneath downtown Los graphically discrete uplift episode as Event W. Using similar observations, we identified www.sciencemag.org SCIENCE VOL 300 4 APRIL 2003 115 three other temporally discrete periods of uplift during the past �11,000 years (Table 1, Events Y, X, and V) (14), as well as evidence for additional older uplift and sedimentary thicken- on October 13, 2008 ing (15). We attribute these four uplift events to coseismic fold growth during large earthquakes Fig. 2. Borehole results from the Carfax Avenue transect. Cross section of major stratigraphic units on the PHT (16), on the basis of the episodic (eight times vertical exaggeration). The thin vertical lines are boreholes. Colors denote different nature of the uplift and the spatial association sedimentary units. Numbers show key units discussed in the text. Thin red lines mark the tops of major of the zone of uplift with the active axial sand- and gravel-filled channels. Double-headed red vertical arrows along the left side of the figure show surface (the locus of active folding) of the fold the stratigraphic range of sedimentary growth after uplift events; green boxes show sections with constant sedimentary thickness across the transect (see text for discussion). Proposed correlations forming above the tip of the deep PHT ramp between Carfax boreholes and a water well (1589S) located 175 m north of the transect are shown by (7, 9, 10). These discoveries obviate concerns dashed stratigraphic contacts along the right side of the figure. Contacts at 12, 17, and 20 m are that near-surface deformation above blind correlative between borehole 22 and the water well. G, ground; S, surface. thrust faults may be too diffuse to record www.sciencemag.org useful paleoseismological information and demonstrate that earthquake ages and displace- Fig. 3. Schematic reconstructions of growth strata showing the kinematic ments can be extracted from blind thrust faults relation between development of fold with precision comparable to that of tradition- scarps and slip on the PHT. Colors of al paleoseismologic studies. young growth strata are the same as for The well-constrained geometry and kinemat- Carfax Units 30 through 50 shown in Fig. 2. (A) Deposition of Units 47 and 46 ics of the PHT and its associated folds (7, 9, 10) Downloaded from allow us to convert uplift in Events V through Y (and probably the lower part of 45) at a near-horizontal stream gradient after to estimates of coseismic displacement on the burial of the Event V fold scarp by Unit PHT. We have conservatively assumed simple 50 sand. (B) Coseismic development of rigid-block motion of the hanging wall up the a 2-m–high fold scarp in earthquake W. PHT ramp (i.e., that no strain is consumed by (C) Onlap of the Event W fold scarp by deformation of the hanging wall block), yielding Unit 40 sand. (D) Deposition of Units minimum measurements of reverse slip on the 30 to 35 at a near-horizontal stream gradient after Event W. The steeply dip- PHT ramp in past earthquakes (Table 1). The ping black dashed line shows the active minimum 13.7 � 1.7 m of slip in these earth- anticlinal axial surface; the gray dashed quakes indicates a minimum slip rate of 1.1 to line shows the inactive synclinal axial 1.6 mm/year since the 9.5 to 10.7 ka age of surface (7, 9, 10). Event V. We estimate a conservative range of paleomagnitudes for these earthquakes by as- suming that our minimum slip estimates represent either average slip or maximum slip during past PHT ruptures (17). Comparison with re- gressions of maximum and average slip versus Mw (18) indicates that Events Y and X were on the order of Mw 7.2 � 0.2/–0.4, whereas Event W was Mw �7.5 � 0.3/–0.6 and Event V was Mw �7.4 � 0.3/–0.4 (19).

www.sciencemag.org SCIENCE VOL 300 4 APRIL 2003 117

405 600 400000 420000

San Gabriel Fault 500 N

5 km 3780000 15 LA California Shaw et al. (2002)

900 B2 (A) Map of three main structural segments of the Pacific Ocean 15 15 10 km Puente Hills Thrust (PHT). From west to east downtown 1987 M6.0 these are the Los Angeles segment, the Santa Fe Los Angeles Springs segment, and the Coyote Hills segment, which itself comprises two distinct sub-segments. B1 10 10 km (B) Cross section based on petroleum-industry

5 3760000 34° seismic relfection data (to 7 km depth) and mainshock and relocated aftershocks of the 1987 PHT Mw 6.0 Whittier Narrows earthquake (Shaw & 5 5 Shearer, 1999 Science) from 11–15 km depth). 5 Brea

= upper tip line of PHT ramp B 118°

B B1 2 km B2 Qt S.l.

Tfu A T

Tfl WF Puente Hills5 Blind-Thrust System, Los Angeles, California 2959

MF (km) depth Tp

middle Miocene PHT Davis, T. L., J. Namson, and R. F. Yerkes (1989). A cross section of the & older Los Angeles area: seismically active fold-and-thrust belt, the 1987 10 Whittier Narrows Earthquake and earthquake hazard, J. Geophys. Res. 94, 9644–9664. Dolan, J. F., K. Sieh, T. K. Rockwell, R. S. Yeats, J. Shaw, J. Suppe, G. 1987 M6.0 Huftile, and E. Gath (1995). Prospects for larger or more frequent 15 earthquakes in the greater Los Angeles metropolitan region, Califor- nia, Science 267, 199–205. Dolan, J. F., K. Sieh, T. K. Rockwell, P. Guptill, and G. Miller (1997). Active tectonics, paleoseismology, and seismic hazards of the Hol- lywood fault, northern Los Angeles basin, California, Geol. Soc. Am. Bull. 109, 1595–1616. Dolan, J. F., K. Sieh, and T. K. Rockwell, (2000). Late Quaternary activity and seismic potential of the Santa Monica fault system, Los Angeles, California, Geol. Soc. Am. Bull. 112, 1559–1581. Eguchi, R. T., J. D. Goltz, C. E. Taylor, S. E. Chang, P. J. Flores, L. A. Johnson, H. A. Seligson, and N. C. Blais (1998). Direct economic losses in the Northridge earthquake: a three-year post-event perspective, Earthquake Spectra 14, 245–264. Erslev, E. A. (1991). Trishear fault-propagation folding, Geology 19, 617– Figure 14. Perspective view of a three-dimen- 620. sional model describing the PHT in relation to other Hauksson, E. (1990). Earthquakes, faulting and stress in the Los Angeles major blind-thrust systems in the northern Los An- Basin, J. Geophys. Res. 95, 15365–15394. geles basin. PHT segments: CH, Coyote Hills; SFS, Hauksson, E., and L. M. Jones (1989). The 1987 Whittier Narrows Earth- Santa Fe Springs; LA, Los Angeles segment; SV, San quake sequence in Los Angeles, southern California: seismological Vicente fault; LC, Las Cienegas fault (Schneider et and tectonic analysis, J. Geophys. Res. 94, 9569–9589. al., 1996); uEP, upper Elysian Park thrust (Oskin et Heaton, T. H., J. F. Hall, D. J. Wald, and M. W. Halling (1995). Response al., 2000); lEP, lower Elysian Park (Davis et al., of high-rise and base-isolated buildings to a hypothetical Mw 7.0 1989; Shaw and Suppe, 1996); CP, Compton ramp blind-thrust earthquake, Science 267, 206–210. (Shaw and Suppe, 1996). King, G. C. P., and C. Vita-Finzi (1981). Active folding in the Algerian earthquake of 10 October 1980, Nature 292, 22–26. Medwedeff, D. A. (1992). Geometry and kinematics of an active, laterally 7.1) earthquakes could occur with repeat times of 780–2600 propagating wedge thrust, Wheeler Ridge, California, in Structural Geology of Fold and Thrust Belts, S. Mitra and G. Fisher (Eds.) John years. As shown by the 1994 Northridge (Mw 6.7) event, Hopkins University Press, Baltimore, 3–28. earthquakes of these magnitudes would cause severe damage Myers, D. J. (2001). Structural geology and dislocation modeling of the in the LA metropolitan area. east Coyote anticline, Eastern Los Angeles basin, M.S. Thesis, Oregon State University, 49 pp. Oskin, M., K Sieh, T. Rockwell, G. Miller, P. Guptill, M. Curtis, S. McAr- Acknowledgments dle, and P. Elliot (2000). Active parasitic folds on the Elysian Park anticline: Implications for seismic hazard in central Los Angeles, This research was supported by the National Earthquake Hazards Re- California, Geol. Soc. Am. Bull. 112, no. 5, 693–707. duction Program (Grant 01HQGR0035), SCEC, and the National Science Pratt, T. L., J. H. Shaw, J. F. Dolan, S. Christofferson, R. A. Williams, J. K. Foundation (EAR-0087648). The authors thank Michelle Cooke, Karl Odum, and A. Plesch (2002). Shallow folding imaged above the Pu- Mueller, Tom Rockwell, and Tom Wright for their helpful insights and ente Hills blind-thrust fault, Los Angeles, California, Geophys. Res. John Suppe and Robert S. Yeats for constructive reviews. Lett. (in review). Schneider, C. L., C. Hummon, R. S. Yeats, and G. Huftile (1996). Structural References evolution of the northern Los Angeles basin, California, based on growth strata, Tectonics 15, 341–355. Allmendinger, R. W. (1998). Inverse and forward numerical modeling of Scientists of the U.S. Geological Survey and the Southern California Earth- trishear fault-propagation folds, Tectonics 17, 640–656. quake Center (1994). The magnitude 6.7 Northridge, California, Allmendinger, R., and J. H. Shaw (2002). Estimation of fault-propagation Earthquake of 17 January 1994, Science 266, 389–397. distance from fold shape: implications for earthquake hazards assess- Shaw, J. H., and P. M. Shearer (1999). An elusive blind-thrust fault beneath ment, Geology 28, no. 12, 1099–1102. metropolitan Los Angeles, Science 283, 1516–1518. Argus, D. F., M. D. Heﬂin, A. Donnellan, F. H. Webb, D. Dong, K. J. Shaw, J. H., and J. Suppe (1994). Active faulting and growth folding in the Hurst, D. C. Jefferson, G. A. Lyzenga, M. M. Watkins, and J. F. eastern Santa Barbara Channel, California, Geol. Soc. Am. Bull. 106, Zumberge (1999). Shortening and thickening of metropolitan Los An- 607–626. geles measured and inferred by using geodesy, Geology 27, 703–706. Shaw J. H., and J. Suppe (1996). Earthquake hazards of active blind-thrust Bawden, G. W., W. Thatcher, R. S. Stein, K. W. Hudnut, and G. Peltzer faults under the central Los Angeles basin, California, J. Geophys. (2001). Tectonic contraction across Los Angeles after removal of Res. 101, 8623–8642. groundwater pumping effects, Nature 412, 812–815. Shearer, P. M. (1997). Improving local earthquake locations using the L1 Barbier, M. G. (1983). The Mini-Sosie Method, International Human Re- norm and waveform cross correlation: Application to the Whittier sources Development Corporation, Boston, 186 pp. Narrows, California, aftershock sequence, J. Geophys. Res. 102, Blake, G. H. (1991). Review of Neogene biostratigraphy and stratigraphy 8269–8283. of the Los Angeles basin and implications for basin evolution, in Stein, R. S., and G. Ekstrom (1992). Seismicity and geometry of a 110-km- Active Margin Basins, K. T. Biddle (Ed.), American Association of long blind thrust fault, 2, Synthesis of the 1982–85 California Earth- Petroleum Geologists Memoir 52, 135–184. quake Sequence, J. Geophys. Res. 97, 4865–4883. JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 114, B12305, doi:10.1029/2008JB006129, 2009 Click Here for Full Article

Evidence for large Holocene earthquakes on the Compton thrust fault, Los Angeles, California Lorraine A. Leon,1 James F. Dolan,1 John H. Shaw,2 and Thomas L. Pratt3 Received 30 September 2008; revised 28 May 2009; accepted 21 July 2009; published 16 December 2009.

[1] We demonstrate that the Compton blind thrust fault is active and has generated at least six large-magnitude earthquakes (Mw 7.0–7.4) during the past 14,000 years. This large, concealed fault underlies the Los Angeles metropolitan area and thus poses one of the largest deterministic seismic risk in the United States. We employ a methodology that uses a combination of high-resolution seismic reflection profiles and borehole excavations to link blind faulting at seismogenic depths directly to near-surface fault-related folding. Deformed Holocene strata record recent activity on the Compton thrust and are marked by discrete sequences that thicken repeatedly across a series of buried fold scarps. We interpret the intervals of growth as occurring after the formation of now-buried paleofold scarps that formed during uplift events on the underlying Compton thrust ramp. Minimum uplift in each of the scarp-forming events, which occurred at 0.7–1.75 ka (event 1), 0.7–3.4 ka or 1.9–3.4 (event 2), 5.6–7.2 ka (event 3), 5.4–8.4 ka (event 4), 10.3–12.5 ka (event 5), and 10.3–13.7 ka (event 6), ranged from 0.6 to 1.9 m, indicating minimum thrust displacements of 1.3 to 4.2 m. Such large displacements� � are consistent with the � occurrence of large-magnitude earthquakes (Mw 7). This multidisciplinary methodology provides a means of defining the recent seismic� behavior, and therefore the hazard, for blind thrust faults that underlie other major metropolitan regions around the world. Citation: Leon, L. A., J. F. Dolan, J. H. Shaw, and T. L. Pratt (2009), Evidence for large Holocene earthquakes on the Compton thrust fault, Los Angeles, California, J. Geophys. Res., 114, B12305, doi:10.1029/2008JB006129.

1. Introduction has been documented [King and Vita-Finzi, 1981; Stein and PHT SFS King, 1984; Stein and Yeats, 1989; Lin and Stein, 1989; Lai [2] The recent occurrence of several highly destructivesegment et al., 2006; Chen et al., 2007; Streig et al., 2007] and in earthquakes on thrust faults (e.g., 1994 M 6.7 Northridge,study site w very large thrust fault earthquakes, discrete coseismic fold California, 1999 M 7.6 Chi-Chi, Taiwan, 2005 M 7.5 w w scarps have developed above bends and at the tips of Kashmir, Pakistan, and 2008 M 7.9 Wenchuan, China) w N. Comptonpropagating faults [Stein and Yeats, 1989; Ishiyama et al., highlights the need to better understand the seismic hazards LA study2004, site 2007; Lai et al., 2006; Chen et al., 2007; Streig et al., posed by these structures to an increasinglyPHT urbanized 2007]. The youngest folded strata thus contain the informa- global population. The keys to this understandingLA segment are the tion necessary to understand the recent seismic behavior of recognition of the rate at which these faults storestudy and site release Active axial these faults. However, extracting paleoseismological data seismic energy, and the acquisition of detailed paleoseismo- from blind thrustss hasurf provedace to be challenging, although logical data on the ages and displacementsst of ancient ru the development over the past several years of new inter- earthquakes that they have generated.th Although such data s ill disciplinary methodologies for documenting the sizes and are routinely documentedH for many faults around the world, e ages of past events is beginning to yield a new understand- one class of faults, thent so-called ‘‘blind thrust faults,’’ in e ing of the behavior of this problematic class of faults at the which the fault planePu does not extend all the way to the time scales of individual earthquakes [e.g., Dolan et al., Earth’s surface, has proved particularly difficult to study. 2003; Sugiyama et al., 2003; Ishiyama et al., 2004, 2007; [3] Unlike most faults, which rupture to the surface in Leon et al., 2007]. The difficulties in studying the paleo- large earthquakes, near-surface deformation above blind seismology of blind thrusts are particularly well illustrated thrust faults is accommodated by folding, rather than fault- by the Compton thrust, a major blind thrust fault beneath ing. In many thrust fault earthquakes, coseismic fold growth metropolitan Los Angeles. Although this fault was declared

1 to be inactive on the basis of earlier studies [Mueller, 1997], Department of Earth Sciences, University of Southern California, Los ourCo investigationsmpton ram demonstratep that the Compton fault is Angeles, California, USA. 2Department of Earth and Planetary Sciences, Harvard University, indeed active and capable of generating large-magnitude Cambridge, Massachusetts, USA. (Mw 7) earthquakes. In this article we discuss the 3U.S. Geological Survey, University of Washington, Seattle, Washington, implications� of these results for seismic hazard assessment USA. in Los Angeles and describe a standard, multidisciplinary methodology that can be used to define seismic hazards Copyright 2009 by the American Geophysical Union. 0148-0227/09/2008JB006129$09.00

B12305 1 of 14 B12305 LEON ET AL.: HOLOCENE EARTHQUAKES ON COMPTON THRUST B12305

Figure 3. Compton fault seismic reflection data illustrating the multidisciplinary methodology utilized in this study [Pratt et al., 2002; Shaw et al., 2002; Dolan et al., 2003] (see also similar studies by Sugiyama et al. [2003] and Ishiyama et al. [2004, 2007]). (a) Hammer source seismic reflection profile (migrated; 8 vertical exaggeration). The prominent reflector at 10–17 m depth shows the kink band clearly. Thin,� vertical, white lines show borehole locations. Dashed white lines represent active, synclinal and inactive, anticlinal axial surfaces. Black box indicates location of borehole transect cross section in Figure 4. A–A0 marks the trace of the seismic reflection profile in Figure 2. (b) Weight drop source seismic reflection profile (migrated). The prominent reflector at 120–200 m depth shows the kink band well. The subhorizontal reflector at about 100 m depth is likely the water table, as velocities above the reflector are consistent with unsaturated strata. The reflector also cuts across the geologic structure consistently defined by the underlying reflectors and overlying drill hole data, as would be expected for the water table. Thin, vertical, white lines show borehole locations. Dashed white lines represent active, synclinal and inactive, anticlinal axial surfaces. Black box indicates location of hammer source profile shown in part A. B–B0 marks the trace of the seismic reflection profile in Figure 2. (c) Seismic reflection image of the back-limb fold structure showing upward narrowing zone of active folding (growth triangle) delimited by sharply defined axial surfaces [Shaw and Suppe, 1996]. Dashed white lines represent active, synclinal and inactive, anticlinal axial surfaces. Black box indicates location of weight drop source profile shown in part B. C–C0 marks the trace of the seismic reflection profile in Figures 1 and 2. (d) Fault bend fold model of back-limb fold structure of the Compton fault with discrete axial surface yielding an B12305 upward narrowingLEON fold ET AL.:limb inHOLOCENE growth strata EARTHQUAKES (modified after Shaw ON and COMPTON Suppe [1996]). THRUST B12305

petroleum industry seismic reflection profiles provide an located along this trend on a previously marshy, extremely exceptionally clear image of the folding above the Compton low-gradient floodplain, facilitating reconstruction of orig- thrust ramp beneath the site. In particular, these data show inally near-horizontal floodplain surfaces and allowing that the back-limb axial surface, the locus of folding collection of numerous organic-rich samples for radiocar- associated with fault slip across the base of the Compton bon dating. It is also far removed from structural interference ramp at depth, exhibits an upward narrowing growth associated with other active faults, such as the Newport- triangle that extends to within 200 m of the surface [Shaw Inglewood right-lateral strike-slip fault (Figure 2). Twentieth and Suppe, 1996]. This fault� bend fold, referred to as the century topographic maps of the site show the ground Compton–Los Alamitos trend, marks the base and north- surface as near horizontal, with an extremely gentle dip of eastern extent of the Compton ramp. Our study site is 0.2� to the south. However, presumably because of histor-

4 of 14

Figure 4. (a) Borehole results from the Stanford Avenue transect. Cross section of major stratigraphic units (16 vertical exaggeration) showing the details of the most recent uplift event (event 1). Thin vertical lines� are boreholes. Green horizontal line is the present ground surface. (b) Cross section of major stratigraphic units (8 vertical exaggeration). Colors denote different sedimentary units. Thin red lines mark the tops of major� sand and gravel units. Double-headed red vertical arrows along the right side show the stratigraphic ranges of intervals of sedimentary thickening across the transect, with the uplift in each event shown to the right of each arrow. Double-headed green vertical arrows show intervals of no sedimentary growth. ical withdrawal of water and petroleum, the ground surface material1 Figure S1). Grain size was visually identified using now dips slightly (0.06�) to the north. The gentle, near- a handheld GSA grain size card, and we assigned soil colors horizontal regional slope at the study site has resulted in the on wet material using a Munsell color chart. deposition of laterally extensive sedimentary layers that can [10] Deposition of these units in a low-gradient floodplain be correlated over large distances. has resulted in remarkable lateral continuity; more than [8] As is important with all paleoseismic studies, we 25 units are traceable continuously across the entire 760 m chose a site characterized by continuous, relatively rapid length of the borehole transect. Calibrated radiocarbon dates sediment accumulation, leading to stratigraphic separation from 30 humic, charcoal, and bulk soil samples indicate that of event horizons and preservation of event stratigraphy. We sediment accumulation has been relatively continuous at specifically chose a study site in which the uplifted hanging 2 mm/yr over the past 14,000 years (Figure 5). The wall exhibited no topographic expression in order to ensure sediment� accumulation reveals no major hiatuses and the that our site was located in an area of continuous aggradation. few soils observed in the section consist of weakly devel- The random occurrence within any floodplain of topographic oped A–C profiles that record brief periods of pedogenesis. features would serve to decrease the consistency of strati- We see exceptionally well defined growth in exactly the graphic growth patterns. Moreover, historical evidence same location throughout the stratigraphic section. This reveals that the floodplain was so low gradient, that during argues strongly that we are seeing a tectonic signal and that rainy winters, the entire Los Angeles Basin (>20 km wide) any natural variation of the floodplain is not a significant was a marsh or shallow lake with standing surface water source of error. We can also directly demonstrate through [Gumprecht, 2001; Linton, 2005]. overlapping radiocarbon dates that at least some of these [9] We excavated a north-south transect of ten 25 to 35 m units are synchronous along the length of the transect. For deep, continuously cored boreholes across the zone of example, three dates from the shallowest of several con- young folding imaged in the high-resolution seismic reflec- tinuous marsh deposits (unit 11, organic rich soil beneath tion profiles. Excavation of boreholes allows us to docu- unit 10) yield overlapping radiocarbon dates that confirm ment the details of the most recent folding events, as well as simultaneous development across the transect of this layer. to sample the sediments for age control. The cores reveal sequences of fluvial sand and gravel, interbedded with cohesive overbank silt and clay layers (Figure 4; auxiliary 1Auxiliary materials are available in the HTML. doi:10.1029/ 2008JB006129.

5 of 14